Re: [softrock40] Help With U5 Diagnosis (SoftRock Lite)
Hi Bill,There have been a few cases where U5 in the SoftRock Lite has been damaged when plugging the SoftRock's audio cable into a soundcard. I think the damage has been caused by an AC potential difference between the SoftRock circuit ground and the soundcard circuit ground. Pin 1 of U5 is connected to the tip of the audio cable so when plugging the cable into the soundcard the tip first contacts the soundcard ground of the line-in jack.
Would you be able to make a voltage measurement between the SoftRock circuit ground and the ground of your soundcard, (barrel of line-in jack), when the SoftRock is connected to your power supply and without the audio cable plugged into your soundcard?
Thanks and 73,
Tony KB9YIG----- Original Message -----From: KR8L/9Sent: Sunday, February 03, 2008 6:34 PMSubject: [softrock40] Help With U5 Diagnosis (SoftRock Lite)
I'm hoping someone can take a look at my analysis and tell me if my
conclusion is correct.
I just completed a SoftRock Lite for 20m. Although I would consider
myself a fairly experienced builder, I have never done any surface
mount work. Thanks to some advice from one of the local hams I jumped
right in, and was very pleased with my results. I used my Weller
Portasol butane torch with hot air tip for the SMD work, and the same
iron with a 1mm tip for the leaded components. It took me several
hours because I work slowly and carefully and inspect my work
frequently with both a 12x loupe and a 4x magnifier (as well as
working under a large lighted magnifier). I had a blast!
After hooking up the little rig and doing some listening with Rocky, I
noticed that I had the image problem that so many have posted about.
I searched the forum, then started troubleshooting. Using Audacity
(http://audacity. sourceforge. net) to monitor the output, I quickly
determined that I had no "I" output.
I started checking the circuit starting with the 3.5mm plug that
connects to my sound card and working back. I checked continuity and
component values back to the left side of U5, then traced back to U4.
I checked all connections both on the bottom of the board and from
the bottom to the top, and verified all resistance readings. Any
capacitor that was not paralleled with a resistor was checked using my
DMM. I measured in such a way that I was verifying continuity through
all solder joints, not just continuity to the pad itself. So far I
had turned up nothing that looked wrong.
In order to test T1 I disconnected a probe from my DMM and plugged it
into the antenna jack of my FT-817, which was tuned to 20m. I touched
the T1 primary then tuned in a CW signal. Then I touched U4 Pins 6,
3, 11, and 12 in turn and verified that the signal strength appeared
to be equal on all pins which I took to be an indication that T1 is
wired and connected correctly. (I'm no stranger to toroidal
transformers and inductors, by the way.)
Next, I checked voltages, in particular making comparisons between the
corresponding pins in the I and Q portions of the circuit on U4 and
U5. (For example, U4 Pin 7 corresponds to U4 Pin 9; U5 Pin 2
corresponds to U5 Pin 6, etc.) Everything matched until I got to U5
Pins 1 and 7. Pin 7, which is in the Q circuit (and which appears to
be working OK) was at 2.37 volts, while Pin 1 (the corresponding pin
in the I circuit) was at 2.71 volts.
Finally, I tried the "scratch" or "hum" test. While watching the
output on the Audacity display and listening to my computer with the
sound card input patched into the output, I touched U5 Pin 6, which is
the input to the Q audio amplifier, with my DMM probe. As expected, I
got a significant hum. When I touched Pin 2, which is the
corresponding pin in the Q circuit I heard nothing. No amount of
scratching or poking could produce any noise in the output.
I don't have anything sensitive enough to measure the output from U4
Pins 7 and 9, but I think my testing does a reasonable job of making
me suspect that the left side of U5 is defective. I was careful to
use electrostatic discharge precautions throughout construction, but
maybe I "zapped" it anyway, or maybe it was defective to start with.
So, do my conclusions sound reasonable? Anything else I should be
Thanks for any help or comments,
73 de Bill, KR8L/9
- --- In email@example.com, FRANCIS CARCIA <carcia@...> wrote:
>into a mixer Rohde put in "Ham Radio" back in '77 or so. 4 - 2N5109s
> Hi All,
> I was going through old notebook #1 of cool RX circuits and ran
referenced to ground. Sure looks like an H mode mixer to me. Not
quite as old as the 4 tube version I found in the West Coast
Handbook. frank WA1GFZ
>Hi Francis and All,
Old book are very important to renew ideas and to learn ...
The double balanced mixer you are referring to is not an H-Mode
Mixer. The circuit diagram may confuse you because it may looks
similar to my 2 Transformer version of the H-Mode Mixer, having, in
particular, a 5 windings transformer.
Ulrich's design is an active mixer, using 4 transistors 2N5109, and
also an intelligent application of the classic passive double
balanced 4 diodes mixer. The circuit has the LO and RF input
from/through the same transformer (5 windings) driving each
transistor base with both signals (RF and LO). The transistors are
doing the mixing work and the output transformer is working as a
Ulrich's, DJ2LR/N1UL, key point in this circuit is the application of
rf feedback to each transistor stage obtaining superior IMD
performance versus prior fet and tube mixer designs.
Ed Oxner, KB6QJ, worked on mixers since the 70s while at Siliconix,
going from FET, transistors to power mosfets. In the 80s he came out
with the double balanced mixer using the Siliconicx Si8901 monolithic
quad ring mosfet and later a commutating mixer with resonant drive
transformer. It was the best mixer at that time.
In 1993 Jacob Makhinson, N6NWP, presented a High Dynamic HF front-end
based on the Si8901, driven by a 74HC74 to provide a square wave LO
N6NWP circuit was modified by Colin Horrabin, G3SBI, using the SD5000
FET array and a 74AC74 to provide square-wave injection. A further
step was the development of the well known H-Mode Mixer using 3
transformers. What makes the H-Mode Mixer different from all others
is not the circuit diagram "form" looking like an H but moving the
mixing action from a semiconductor to the transformers, thanks to the
FETs switching to ground the output transformers windings. This also
brought the benefit of not having the LO modulating the RF signal as
in the previous configurations where the FETs or transistors were in
serial between transformers windings. In 1998 the SD5000 was
replaced by the FST3125.
Yaesu, lately Icom and now Elecraft K3 are using the "basic" Oxner
mixer configuration with mosfet switches array. K3 has not an H-Mode